基于改进模糊C均值聚类的弥散张量成像图像分割  

DTI image segmentation based on the improved fuzzy C-means clustering

作　　者：房博文 张相芬[1] 马燕[1] 李传江[1] 张玉萍[1] 杨燕勤[1] Fang Bowen;Zhang Xiangfen;Ma Yan;Li Chuanjiang;Zhang Yuping;Yang Yanqin(The College of Information,Mechanical and Electrical Engineering,Shanghai Normal University,Shanghai 200234,China)

机构地区：[1]上海师范大学信息与机电工程学院,上海200234

出　　处：《上海师范大学学报（自然科学版）》2018年第4期474-478,共5页Journal of Shanghai Normal University(Natural Sciences)

基　　金：国家自然科学基金(61373004);上海师范大学校级基金(A700115001005;Sk201220)

摘　　要：为了避免随机选取初始聚类中心点的缺陷,利用最大最小距离的方法确定初始聚类中心点.实验结果表明,和传统的模糊C均值聚类(FCM)算法相比,所提聚类算法具有较高的稳定性和准确性,所分割的胼胝体图像边缘信息更加清晰.In order to avoid the shortcomings of selecting initial clustering center points randomly,we use the principle of maximum and minimum distance to determine the initial clustering center points. Compared with traditional fuzzy C-means clustering（FCM）algorithm. The experimental results show that the accuracy and stability of improved FCM algorithm has been improved,and corpus callosum edge information is clearer.

关 键 词：模糊C均值聚类(FCM) 最大最小距离 初始聚类中心 胼胝体 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]